Biomimetic underwater robots driven by flapping wing propulsion exhibit strong nonlinear coupling between surge and heave, complicating motion control. Linear controllers are effective near nominal conditions but degrade under nonlinear dynamics, while the effectiveness of adaptive methods such as
MRAC remains uncertain. This paper compares a parallel linear controller and a MIMO MRAC scheme using a nonlinear robotic penguin model with hydrodynamic effects. Results show that the parallel controller
achieves stable tracking with low overshoot (< 8%) and lower RMSE, whereas MRAC provides faster rise times (∼0.5–1.2 s) but suffers from large overshoot (up to 25%) and oscillatory behavior. These findings indicate that the parallel approach offers a more reliable and practical solution, while MRAC requires further tuning to ensure stability under strong coupling.
ヘルプ